Google’s Gemini AI embeds extractive techno-optimism: 3D simulations deepen corporate control over knowledge while obscuring material costs of AI infrastructure

Indigenous knowledge systems emphasize relationality, place, and embodied learning—values fundamentally at odds with AI-generated simulations that abstract reality into manipulable, decontextualized models. The technology’s reliance on Cartesian spatialization erases Indigenous cosmologies that see land, water, and sky as kin rather than data points. Moreover, the extraction of rare earth minerals for servers mirrors historical colonial resource plunder, deepening cycles of dispossession.

This innovation echoes earlier waves of technological enclosure, from the printing press’s role in standardizing knowledge to television’s centralization of cultural narratives. Each era’s ‘disruptive’ tool was framed as progress while deepening corporate control over information. The current AI boom repeats the pattern: proprietary platforms absorb public knowledge, monetize it, and sell it back as ‘personalized’ convenience, mirroring the enclosure of the commons in the 18th century.

In contrast to Western computational epistemology, many African and Indigenous systems view knowledge as fluid, communal, and tied to ancestral wisdom. The AI’s 3D simulations impose a static, user-driven framework that fragments collective understanding into individualized interactions. This risks reinforcing colonial hierarchies of knowledge, where Western tech defines ‘truth’ while marginalizing oral, land-based, and spiritual ways of knowing.

The energy demands of generating 3D models and real-time simulations are substantial, with AI models consuming up to 10x more compute than text-only outputs. Google’s claims of ‘efficiency’ obscure the carbon footprint of data centers, which now rival some nations’ electricity consumption. Scientifically, the technology’s ‘accuracy’ is contingent on training data biases, which disproportionately reflect Western, urban, and male perspectives.

Artistic traditions often critique the reduction of reality to manipulable objects, as seen in surrealism’s rejection of rationalist frameworks or Indigenous art’s refusal to separate form from meaning. The AI’s 3D models flatten complexity into interactive toys, mirroring the spiritual emptiness of consumerist progress narratives. Meanwhile, spiritual practices like meditation emphasize presence over simulation, highlighting the dissonance between AI’s hyper-stimulation and contemplative wisdom.

By 2030, AI-generated simulations could dominate education, urban planning, and medicine, creating feedback loops where corporate-controlled models shape public perception of reality. This risks a future where only those with access to premium AI tools can ‘see’ the world accurately, deepening digital divides. Scenario planning must account for the geopolitical tensions over data sovereignty and the potential collapse of shared epistemic frameworks.

Marginalized communities—especially in the Global South—face the dual burden of bearing the environmental harms of AI infrastructure while being excluded from its design and benefits. Data laborers in Kenya, India, and the Philippines train AI models for poverty wages, yet their knowledge is erased in the final product. The technology’s Western-centric interface further alienates non-English speakers and those without access to high-speed internet.